Nanofiltration aids catalyst recovery

The recovery of transition metal catalysts using organic solvent nanofiltration has commercial and environmental benefits, as Dr Ali Nozari, business development manager at Membrane Extraction Technology (MET) explains

The development and use of transition metal catalysts has greatly benefited the synthesis of novel organic compounds over the past 50 years. As the complexity of the target product compounds has increased, so has the design of transition metal catalysts (TMCs). Catalysts can now aid a wide range of reactions from relatively simple carbon-carbon bond formation and hydrogenation catalysts (eg, Suzuki and Wilkinson's catalysts respectively) to more exotic asymmetric catalysts (eg, Katsuki-Sharpless Ti-based epoxidation catalysts) that utilise chiral ligands to direct the chirality of the final product. The development of many new fine chemicals and active pharmaceutical ingredients (APIs) has resulted in many syntheses requiring expensive TMCs. However, to make these catalytic processes economically viable the catalyst needs to be separated in its active form and reused.

Few industrial processes exist that can recover the catalyst in an active form; instead they focus on obtaining pure, metal-free product by removing the metal catalyst and decomposed TMC fragments. These processes are often energy intensive, generate appreciable amounts of downstream waste, and lead to poor process economics as the ligands, which are often more expensive than the metal itself, are lost.

For the past 10 - 15 years there has been consid-erable academic activity in developing organic solvent nanofiltration (OSN) membranes capable of separating molecular species in the rage 200 - 1000 gmol^-1, and in the past five years an increasing number of OSN membranes (both polymeric and ceramic) have become commercially available.

These commercial OSN membranes increasingly find application in solvent exchange; solvent recovery; product purification, recovery and reuse of catalysts; and the removal of heavy metals (eg, palladium).¹ The development of OSN membranes has made the non-destructive, energy efficient separation and concentration of catalysts from products feasible. Nanofiltration membranes can be used to exploit the molecular weight difference between typical catalysts (ca. > 600 gmol^-1) and reaction products (ca. < 400 gmol^-1) to achieve the desired separation. For example the Starmem range of OSN membranes has been applied to the following separation types:

• Product separation, and catalyst recovery and re-use
• Removal of heavy metals

Post-reaction processing, or work-up, in many catalytic processes can be difficult and complicated, involving many steps, such as washing, salting out, crystallising, adsorbing, etc. OSN solves this problem by providing an efficient route to separating the catalyst from the product, thus allowing the catalyst to be recycled and reused. This can significantly improve the process economics and hence the likelihood of a particular catalyst being used in a process.

MET's catalyst separation technology utilises Organic Solvent Nanofiltration (OSN) as a highly efficient method to separate catalysts.² As well as allowing the catalyst to be re-used in subsequent reactions, a further benefit is that the metal levels in the product stream are much reduced (relative to the post-reaction mixture). This is particularly advantageous in API production. In addition, there are no yield penalties due to adsorption of product when compared with chromatography and adsorption techniques. This enables a reduction in catalyst usage (increased S/C ratio), recovery of high value catalysts and ligands, and a much simpler work-up protocol.

In the case of phase transfer catalysis, the key challenge is to separate the catalyst from the product stream - typically this involves water washes for hydrophilic phase-transfer catalysts, or distillation for hydrophobic catalysts.³ Water washing generates hard-to-treat wastewaters as typical catalysts, such as quaternary amines, are highly toxic to bio-treatment plants, and the catalysts are often surface active, so the generation of emulsions is common. Distillation of hydrophobic catalysts also generates considerable problems due to thermal decomposition of the catalyst. Light decomposition products contaminate the product stream, and heavy decomposition products can form tars and other unwanted distillation residues.

To eliminate these issues, OSN can be applied directly to the product stream once the reaction is complete. The OSN membrane retains the catalyst (which can be recycled and re-used in the reactor) whilst allowing the product to permeate through the membrane with the process solvent. This provides a clean post-reaction stream for additional processing. Example systems for which OSN has been used to recover and reuse catalysts include phase-transfer catalysts (eg, quaternary ammonium catalysts) and organometallic catalysts (eg, chiral hydrogenation catalysts include Pd/Rh complexes, Ru-BINAP, liganded Rh species and Jacobsen catalyst).

MET's catalyst separation technology employed in the Ru-BINAP catalysed hydrogenation of di-methyl itaconate (DMI) has shown that: (i) the reaction can be conducted at a S/C ratio = 500 in the first batch, 1 wt % DMI feed (46 mM); (ii) no further catalyst added after initial batch; no reaction rate decline observed for consecutive 2.5 hour reactions; (iii) constantly greater than 99% conversion of DMI to dimethyl methylsuccinate (DMMS) obtained.

heavy metals removal

One problem with using TMCs, especially when synthesising APIs, is the residual amount of catalyst left in the final product. Another area of novel OSN applications is the use of OSN in combination with other conventional techniques to create a synergistic technology that, in combination, works better than the individual techniques. For example, OSN combined with conventional adsorption techniques (the MemSorb process) can be applied to remove greater than 99% of heavy metals to levels below 10 µg per g product (equivalent to parts per million, ppm) (for palladium), without the yield penalties (due to absorption of product) associated with techniques like chromatography.

For example, MET's MemSorb process was employed in the synthesis of 1-biphenyl-4-yl-ethanone and showed that palladium levels of 2100 ppm (in the post aqueous work-up stream) were reduced to 4 ppm in a single pass of the combined adsorption/OSN process.

As a technology, OSN has the potential to be used in a wide variety of organic synthesis applications and chemical manufacturing processes. OSN can substitute existing operations involving distillation, evaporation or chromatography, with many added benefits compared with conventional techniques. As a technology, OSN is also enabling novel applications to be developed and brought to commercial use that are not feasible using conventional technologies. Also, in light of the Kyoto agreement, it is important to recognise that membrane processes are low energy processes - in comparison to distillation, OSN typically uses less than 10% of the energy for a given separation. Finally, we expect OSN to develop and establish its presence in specialist separations and fine chemicals industry over the coming years.